Current practice is to set up trial crystallization screens and periodically review the results to see if a crystal or promising crystal-like precipitate has appeared a process that often takes weeks or months. Most outcomes are precipitated protein or clear drops, and the conditions that led to those results are removed from further consideration. We are developing an alternative screening approach, the self-association behavior of the target macromolecule as measured by fluorescence anisotropy as a diagnostic for the likelihood of crystallization under the test conditions. Dilute solution properties are known to be a diagnostic for crystallization (George and Wilson, 1994;George et al., 1997;Wilson et al., 1993;Wilson et al, 1996;Tessier et al., 2002;Tessier et al., 2003;Garcia et al., 2003a;Garcia et al., 2003b;Bloustine et al., 2003). Concentration vs. anisotropy data for a macromolecule-precipitant combination was originally proposed for determining the likelihood of that solution producing crystals, although based on Phase I results intensity vs. concentration data is also a strong indicator. Data acquired in Phase I shows that this approach can "find" lead crystallization conditions from solutions that give clear drops or precipitate in screening assays. The applications of the instrument and methodology to be developed will be to rapidly conduct crystallization screens within 2-24 hrs, using a minimum amount of protein ( 0.03 mg), with a higher probability of finding lead conditions. Higher success rates will greatly facilitate structure-based drug design, particularly for target proteins that are difficult to obtain, and contribute to the understanding and treatment human disease. The Phase II proposal's objectives are to substantially improve the present instrument and methodology, progressing from the currently required 4.5 mg to 0.03 mg of protein/96 condition screen, and validate this method with extensive testing. This will be the basis for a macromolecule crystallization business operated on a fee-for-service basis. Experience with the Phase I instrument has indicated where improvements can be made in the plate set-up process, data collection optics, and electronics and the initial Phase II work will be to implement those improvements. Subsequent testing will first be with model proteins. For each model protein the concentration vs. anisotropy and intensity data obtained will be compared with crystallization screens set up in parallel, to define the signature curves indicating crystallization or potential crystallization outcomes and the extended data range over which crystallization conditions can be recovered. All anisotropy-derived leads will be tested with optimization screens. Subsequent testing will be to challenge the methodology using previously uncrystallized soluble and membrane proteins from the same source. The model and test protein data will be used in the subsequent development of software for data analysis. Projected Phase III efforts include developing incomplete factorial screens that can make full use of the quantitative data obtained by this method. PUBLIC HEALTH RELEVANCE: Successful crystallization and X-ray data analysis provides important three-dimensional information on the macromolecules structure-function relationship. Many proteins that are potential drug targets or key components in diseases are only available in trace quantities, or are difficult to obtain. This proposal is to continue development of a new approach to macromolecule crystallization, using a minimum amount of protein, and giving quantitative data that can subsequently be analyzed to determine those conditions which will give crystals and those that can be brought to crystallization conditions, thus giving a higher success rate.